Childhood cancer is the second leading cause of death in children in the United States. Despite significant advances in treatment, its etiology remains enigmatic. Given the well documented role of transplacental carcinogenesis, fetal exposures, particularly maternal use of therapeutic drugs in pregnancy, present an important area for study. In addition to providing etiological clues in the development of cancer, they are avoidable. A major impediment to research in this area is the rarity of the outcomes and the difficulty in obtaining accurate information on exposures remote from the present. We have identified a cohort of approximately 250,000 Tennessee children with 1.7 million person-years of follow-up for whom computerized pharmacy records provide an objective indicator of maternal drug use during pregnancy. The estimated 250 cases of cancer in cohort will be identified from records kept at pediatric hospitals that treat childhood cancer in Tennessee. The primary objectives of this study are to use this cohort to perform a retrospective cohort study of the association of childhood cancer risk with in utero exposure to antihistamines and metronidazole. Both groups of drugs are commonly used during pregnancy, and there is evidence suggesting a link with childhood cancer.